The following is a brief description of the physiological role of Lp(a). The discussion is not meant to be complete and is provided only for understanding of the invention that follows. This summary is not an admission that any of the work described below is prior art to the claimed invention.
Low density lipoproteins (LDLs) are mainly composed of cholesterol, phospholipids and a single hydrophobic protein, apolipoprotein B [apoB]. They are considered as the major carriers of cholesterol in human plasma (for review see Uterman, G. (1989) Science 246, 904-910). ApoB, the only protein subunit of LDL, recognizes and binds to LDL receptors on the surface of cells. This LDL-LDL receptor interaction results in the internalization of the LDL and the eventual release of cholesterol inside the cell.
A modified form of LDL, termed as lipoprotein (a) [Lp(a)], was discovered in 1963 [Berg, K. (1963) Acta Pathol. Microbiol. Scand. 59, 369]. Covalent linkage of an additional glycoprotein, apo(a), to the LDL distinguishes Lp(a) from LDL. Several studies have recently suggested that elevated levels of Lp(a) in human plasma is linked to heart disease (Gurakar, et al., (1985) Atherosclerosis 57, 293-301; Leren, et al., (1988) Atherosclerosis 73, 135-141; Utermann, Supra). The Lp(a) levels range over 1000 fold and individuals with top quartile of plasma Lp(a) levels have two-to five-fold increased probability of developing atherosclerosis.
Atherosclerosis is a disease associated with hardening and loss of elasticity of arterial walls. High concentrations of cholesterol, in the form of Lp(a), in human blood plasma is one of the most important factors responsible for atherosclerosis. Deposition of cholesterol in the Macrophages and smooth muscle cells associated with arterial walls cause plaques (atheromatous lesions) which cause proliferation of adjoining smooth muscle cells. With time, these plaques grow in size causing hardening of the arterial walls and loss of elasticity, which in turn results in rupturing of the arterial walls, blood clotting and blockage of blood flow in the artery (for datails see Textbook of medical physiology Guyton, A. C., (Saunders Company, Philadelphia, 1991) pp. 761-764).
Lp(a) and/or apo(a) levels correlate well with an increased risk of atherosclerosis and its subsequent manifestations such as myocardial infarction, stroke, and restenosis. The apo(a) protein is unique to humans, Old World primates and hedgehogs; its absence in common laboratory animals has made exploration of the physiological role of apo(a) levels difficult. Recently, a transgenic mouse expressing the human gene encoding apo(a) was constructed [Lawn et al., (1992) Nature 360, 670-672]. The transgenic mice are more susceptible than control liter-mates to the development of lipid-rich regions in the aorta. Moreover, human apo(a) expression colocalizes to the regions of fat deposition. Thus, overexpression of apo(a) directly leads to atherosclerotic-like lesions in experimental animals. This observation lends credence to the hypothesis that elevated levels of apo(a) in humans contribute to atherosclerotic disease.
Apolipoprotein(a) is a large glycoprotein which varies in size from 300-800 KDa. Thirty four different isoforms have been characterized from human plasma. The only human cDNA clone currently available encompass 14 kilobase message that encodes apo(a) [McLean et al., (1987) Nature 330, 132-137]. A Rhesus monkey cDNA representing a part of the 3' end of the apo(a) mRNA has also been cloned and sequenced (Tomlinson et al., 1989 J. Biol., Chem. 264, 5957-5965). Sequence analysis of the cloned cDNA revealed two unique facets of the apo(a) structure. First, the apo(a) cDNA is remarkably repetitious. The reconstructed apo(a) cDNA encodes a protein of 4,529 amino acids; 4,210 of the residues are present in 37 repeats of 114 amino acids each. The repeated units themselves are especially homologous; 24 are identical in nucleotide sequence, four more share a sequence that differs in only three nucleotides and the remaining repeats differ by only 11 to 71 bases.
Secondly, apolipoprotein(a) is highly homologous to the serine protease, plasminogen. Plasminogen consists of five repeated homologous domains termed kringles (which are approximately 50% homologous in their amino sequences) followed by a trypsin-like protease domain. Kringle IV of plasminogen is very homologous to the 37 repeats of apo(a) [75-85% at the protein level). In addition, the 5' untranslated region, the signal peptide region, kringle V, the protease domain, and the 3' untranslated region of plasminogen are 98%, 100%, 91%, 94% and 87% homologous to apo(a) sequences, respectively. Relative to plasminogen, apo(a) is missing kringles I, II, and III and, as mentioned above, has extensively duplicated kringle IV. Despite the high degree of homology apo(a) cannot be converted into a protease by tissue type plasminogen activator (tPA). This is because of a single amino acid substitution in apo(a) at the site of activation of plasminogen by tPA (Utermann, supra). IN vitro studies have indicated that apo(a) and Lp(a) compete with plasminogen for binding to the plasminogen receptor and fibrin which supports the correlation between high Lp(a) levels and myocardial infraction (Gonzalez-Gronow et al., (1989) Biochemistry 28, 2374-2378; Hajjar et al., (1989) Nature 339, 303-305; Miles et al., (1989) Nature 339, 301-303). Recent in vivo studies in human (Moliterno et al., 1993 Circulation 88, 935-940) and monkey (Williams et al., 1993 Atheroscler. Thromb. 13, 548-554) support a role for Lp(a) in preventing clot lysis.
The extraordinary homology between apo(a) and plasminogen presents several barriers to drug development. Small molecule inhibitors of apo(a) would have to selectively bind apo(a) without negatively impacting plasminogen function. Similarly, antisense approaches will be limited by the overall nucleotide sequence homology between the two genes. Current dietary and drug therapies (Gurakar, et al., supra; Leren et al., supra), with the exception of nicotinic acid, have little or no effect on apo(a) levels.
Applicant now shows that these same limitations are opportunities for ribozyme therapy. The cleavage site specificity of ribozymes allows one to identify ribozyme target sites present in apo(a) mRNA but completely absent in the mRNA of plasminogen. For instance, there are 13 hammerhead cleavage sites present in the highly conserved kringles of apo(a) that are not present in kringle IV of plasminogen. Likewise, the last kringle repeat, protease domain and 3' untranslated region of apo(a) contain 21 hammerhead ribozyme cleavage sites present in apo(a) that are not present in plasminogen. Thus, ribozymes that target apo(a) mRNA represent unique therapeutics and diagnostic tools for the treatment and diagnosis of those at high risk of atherosclerosis.